Before the piloting, the confidence tier was added below the reason tier of each question. Based on the recommendation of Matell and Jacoby (1971), we used a confidence scale, ranging from 1 (Just Guessing) to 6 (Absolutely Confident), for optimal discrimination and reliability.